bmad-architect
Pass
Audited by Gen Agent Trust Hub on Mar 20, 2026
Risk Level: SAFECOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes local shell scripts, specifically
scripts/nfr-checklist.shandscripts/validate-architecture.sh, to generate Non-Functional Requirements (NFR) checklists and validate architectural documents. - [PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection by ingesting untrusted data from external project documentation.
- Ingestion points: The skill reads content from
docs/bmad/prd.md,docs/bmad/tech-spec.md, anddocs/bmad/architecture.mdto inform its architectural decisions. - Boundary markers: No delimiters or instructions to ignore embedded commands within input data were identified.
- Capability inventory: The skill can write files to the repository and execute bash scripts.
- Sanitization: No sanitization, validation, or escaping of the input document content is performed before the data is processed by the model.
Audit Metadata